Aims

1. To estimate the annual incidence rate of dementia and its subtypes, by age group, education and centre
2. to investigate risk factors for incident dementia and AD, testing the following specific hypotheses:
   1. cardiovascular risk factors (diabetes, hypertension, large waist circumference, high waist hip ratio, hypercholesterolaemia, hypertriglyceridaemia, metabolic syndrome and smoking) and stroke increase the risk for incident dementia and AD, after controlling for confounding effects.
   2. micronutrient deficiencies (vitamin B12 and folate), and anaemia increase the risk for incident dementia and AD, and mediate any effect on these outcomes of overall nutritional deficiency.
   3. sub-clinical hypothyroidism is associated with increased risk of dementia and AD.
   4. incident dementia and AD are inversely associated with the proportion of the genome that is of African ancestry, and the association between APOE e4 genotype and incident dementia and AD is modified by admixture proportions - weaker in those with least African ancestry.
3. To confirm the predictive validity of the survey dementia diagnoses (DSM IV and 'the 10/66 dementia algorithm') and Mild Cognitive Impairment (MCI) through three year follow-up of all dementia and MCI cases
4. To carry out a longitudinal study of evolving care arrangements for people with dementia, and caregiver strain
5. To estimate annual all-cause and cause-specific mortality
6. To investigate the independent effects upon all-cause and cardiovascular mortality of education, socio-economic status, cardiovascular risk factors (CVRF), prevalent cardiovascular disease (CVD), dementia, cognitive impairment, and depression
7. To estimate, the annual attack rates for stroke
8. To compare the risk factor profile for incident stroke (relative risks and population attributable risk fractions) between centres and within centres between men and women and better and less well educated participants